In the Productivity Commission · Freehill Hollingdale & Page SYDCC\SUB34.DOC25 September 1998...

$: In the Productivity Commission · Freehill Hollingdale & Page SYDCC\SUB34.DOC25 September 1998 (15:46) page 2 Proper procedures must be followed by the Productivity Commission This$
Freehill Hollingdale & Page SYDCC\SUB34.DOC25 September 1998 (15:46) page 1

In the Productivity Commission

Inquiry into Pig and Pigmeat Industries: SafeguardAction Against Imports

Submission of the Government of Canada

Introduction

The Government of Canada now submits its further submission to the Productivity Commission for the

purposes of this Inquiry. In so doing, it repeats and reiterates its earlier submission, 1 which should be read

together with this submission.

Canadian pork exports to Australia face yet another inquiry

This inquiry is the latest in a number of Australian government inquiries related to the Australian pig

industry in general and to Canadian imports of frozen processed pork in particular.

In this context we note, in particular, reports concerning the restrictive sanitary limitations on imports of

Canadian pork; Australian Customs Service 2 and Anti-Dumping Authority 3 reports into anti-dumping and

subsidy complaints against Canadian pork; and the Industry Commission’s research project report concerning

pigs and pigmeat.4

Canada’s pork trade with Australia has come under intense scrutiny through these inquiries and reports. At

no stage in the past have the Australian authorities discerned any predatory conduct on the part of Canadian

exporters, nor any causal link between Canadian imports and material or serious injury allegedly suffered by

the relevant Australian industry.

The domestic pig growing industry has also conducted “buy Australian” campaigns against Canadian pork,

including aggressive criticism of major retailers5 and numerous allegations of inaccurate labelling by

Australian smallgoods processors using Canadian pork. 6 In Canada’s view, the industry has also misinformed

the public about imported Canadian pork.7

1 Submitted under cover of letter from the Canadian High Commission dated 13 August 1998.2 Frozen pork from Canada (Report and Preliminary Finding No 92/20, 27 November 1992).3 Review of the Australian Customs Service negative preliminary finding on frozen pork from Canada(Report No. 90, January 1993).4 Pigs and Pigmeat, Industry Commission Research Project, 30 October 1995.5 For example, in a media release dated 16 July 1998, the Pork Council of Australia (“the PCA”)attacked Woolworths for accepting processed products containing imported pigmeat. In part, the President ofthe PCA said “Will Woolworths be selling imported chicken, salmon or beef products if they becomeavailable - regardless of domestic supply?”6 For example, in a media release dated 3 December 1996, the PCA accused Australian smallgoodsmanufacturers of “possible breaches of the Trade Practices Act”. Two months earlier, the PCA in a mediarelease dated 28 October 1996 complained about an alleged “absence of effective country of origin labellinglaws in Australia”.7 For example, in a media release dated 29 May 1998, the PCA claimed that Canadian pork was“subsidised by 15%”. US administrative reviews of alleged subsidies have led to a zero countervailing dutyrate: see Section 5 of this submission. Also, the Australian Pork Corporation submission to the ProductivityCommission refers to an “11,000 metric tonne plateau” of Canadian pork imports, whereas ABS statisticsshow that the 9,000 metric tonne volume for calender 1997, and that 9,000 metric tonne level has not beenmatched (on a rolling 12 month basis) in any month since then.


Proper procedures must be followed by the Productivity Commission

This “safeguards inquiry” was commenced by way of Terms of Reference dated 26 June 1998. The “general

procedures” for inquiries of this sort are contained in a Gazette notice of the previous day. 8

Canada is concerned to ensure that due process is accorded to it as an interested party and member of the

WTO, and that the requirements of the Agreement on Safeguards, both as to procedure and substantive

issues, are carefully complied with. The degree of care required is increased by reason of a number of factors

which are peculiar to this case.

First, there does not appear to be any established legal or administrative framework in Australia for

conducting the inquiry. We make this submission with all due respect for the Productivity Commission which

has been charged by the Australian Government with responsibility for the inquiry. The jurisdiction in

Australia for conducting an inquiry under the Agreement on Safeguards is entirely new. The “temporary

assistance” jurisdiction previously handled by the Industries Assistance Commission and the Temporary

Assistance Authority prior to 1989 was prior to the Agreement on Safeguards, and the temporary assistance

reports are vastly different in terms of criteria and industry policy. 9

Secondly, the status of the Gazette notice is unclear. It is an incomplete rendition of the Agreement on

Safeguards itself, although it seeks to incorporate that Agreement by reference to it. The Gazette notice does

not have statutory legal status, although as a matter of administrative law a failure to comply with it may base

a claim for prerogative relief or relief under the Administrative Decisions (Judicial Review) Act 1977.

Canada trusts and expects that the entirety of the Agreement on Safeguards will be adopted and applied,

notwithstanding the lack of implementation of it as a domestic Australian law and the incomplete replication

of it by way of public gazettal.

Thirdly, Canada has indicated to the Productivity Commission the importance attached to the principle that

proper procedures must be in place to enable interested parties to present evidence and their views, including

the opportunity to respond to the presentations of other parties. Canada has some misgivings about the

adequacy of the opportunity afforded to interested parties to respond, if seen as necessary, to submissions

which may be lodged later in the inquiry period. The opportunity afforded to one or more interested parties to

present a public submission after the date originally notified as being the due date for submissions was not

advised as part of the investigation procedures at the commencement of the inquiry, and impacts on the time

within which responses can be made and fairly considered by the Productivity Commission.10

The keypoints of this submission

Safeguard action under Article XIX of the GATT 199411 allows a Member to derogate from

the fundamental tariff binding commitments under the GATT 1994. Accordingly, the

imposition of safeguard measures is a severe step. The definitional and threshold tests which

must be met to justify such action, and the investigative procedures, specified under the

8 Special Gazette No S297, 25 June 1998.9 The Temporary Assistance Authority was abolished pursuant to the Industries CommissionAmendment Act 1984, and the temporary assistance provisions which were transferred to the IndustriesCommission were repealed by the Industry Commission Act 1990.10 The Commission said, in Sydney on 20 August 1998, that there would be “no further hearings to getelaboration” on written submissions (Transcript at page 158).


Agreement on Safeguards must be strictly complied with. Adequate trade compensation must

be offered to a Member adversely affected.

In considering the imposition of safeguard measures:

“increased imports” of a particular product into Australia must be found to have occurred;

that increase must have been such as to “cause or threaten to cause serious injury” to an

Australian industry;

the Australian industry concerned is that which “produces like or directly competitive

products”;

“objective evidence” must “demonstrate” the existence of this causal link between the

increased imports and the serious injury, or the threat of that injury;

that serious injury must be of a magnitude to have caused a “significant overall impairment

in the position of [the] domestic industry”;

injury caused by “factors other than increased imports” must “not be attributed to

increased imports”;

the relief from the effect of imports must have the character of “emergency action” to

remedy “unforeseen developments”; and

it must be considered “whether or not the application of a safeguard measure would be in

the public interest”.

Canada believes that the Australian industry producing the “like or directly competitive

product” has not been injured seriously, if at all, by Canadian imports. Certainly any injury

falls well short of the “significant overall impairment” required under the Agreement on

Safeguards. This belief is based on a comparison of the condition of the Australian industry,

and the factors influencing its performance, with the requirements for establishing the

relevant degree of injury under the Agreement on Safeguards. The understanding of the

Government of Canada in respect of these matters is based on information provided to us

about the Australian industry by our own agricultural authorities, and by the Canadian pork

and pig growing industries with whom we have consulted for the purposes of providing this

submission to the Productivity Commission.12

Agreement on Safeguards issues addressed by this submission

11 Marrakesh Agreement Establishing the World Trade Organisation.12 The Canadian pork and pig growing industries have advised the Government of Canada that theywill be presenting information to the Productivity Commission on these factual issues by way of separatesubmission.


We refer the Commission to the submission of the Government of Canada lodged on 13 August 1998. We

now provide further submissions for the assistance of the Commission in support of the primary submissions

set out in that previous submission, namely:

• the correct approach to adopt to the question of determining the identity of the

Australian industry producing “like or directly competitive products”;13

• the seriousness of the degree of injury mandated by the Agreement on Safeguards,

and how that injury (“a significant overall impairment in the position of a domestic

industry”) must be considered;14

• the methodology for evaluating whether there is a causal link between increased

imports and the “serious injury” (if any is detected);15 and

• the way the question of “threat” of serious injury should be approached.16

Additionally, this submission seeks to assist the Productivity Commission in its resolution of a number of

specific matters identified by the Commissioner during the public hearings, namely:

• the interrelationship between GATT 1994 Article XIX and the Agreement on

Safeguards itself, in particular in relation to the “emergency” character of relief,

“unforeseen developments” and the meaning of the word “industry” ;17

• the degree of increase in imports needed to justify any consideration of the

imposition of safeguards measures;18 and

• anti-dumping and countervailing. 19

Lastly, this submission comments on and/or rebuts information contained in an economic analysis presented

to the Productivity Commission during the course of the public hearing held on 19 August 1998 in

Brisbane.20

What is the domestic industry producing like or directly competitive products?

The Terms of Reference refer to a specific category of goods

The Terms of Reference refer to specific imported goods, and not others. The relevant tariff item, 0203.29,

can be expressed as follows:

“Frozen meat of swine, not being carcasses or half carcasses, nor hams, shoulders

and shoulder cuts with bones in”

Essentially, the item covers boneless frozen pork. Under the Agreement on Safeguards, it is the domestic

industry producing goods which are “like or directly competitive products” to the relevant imports which

must be the focus of consideration.

13 Section 2 of this submission.14 Section 3 of this submission.15 Section 4 of this submission.16 Section 5 of this submission.17 Section 6 of this submission.18 Section 3.1 of this submission.19 Section 7.20 Section 8 and Attachment 1.


What are the “like goods” to imported boneless frozen pork?

The concept of “like goods” is not defined within the Agreement on Safeguards itself. The words are defined

in the Agreement on the Implementation of Article VI of GATT 1994 (“the Anti-Dumping Agreement”), and

despite the different context it is worthwhile to explain how those words have been defined in another part of

the same international agreement, and also how the relevant Australian authorities have considered those

words in that context.21

Under Article 2.6 of the Anti-Dumping Code:

“ ‘like product’... [means] a product which is identical, ie. alike in all respects to

the product under consideration, or in the absence of such a product, another

product which although not alike in all respects, has characteristics closely

resembling those of the product under consideration.”

In Review of the Australian Customs Service negative preliminary finding on frozen pork from Canada,22 the

Anti-Dumping Authority was faced with the question of extending the definition of the industry, for material

injury assessment purposes, under the “close processed agricultural goods” provisions of Section 269T(4A)

and (4B) of the Customs Act 1901. That was a decision made in a very different context to the “directly

competitive” criteria under the Agreement on Safeguards. In respect of the “like goods” issue both Customs

and the Anti-Dumping Authority focussed on pork cuts, and not on pig carcasses or pigs themselves. The

Anti-Dumping Authority concluded that “the frozen pork exported to Australia from Canada was a like good

to both pork sold within Canada and pork produced within Australia”.

We submit that this remains the proper approach, namely that the industry performing the activity of pork

processing is the industry producing “like goods” which must be assessed by the Commission to determine

whether the injurious effects of imported pork have been so serious as to justify the application of safeguard

measures. It is submitted that there can be no real contest in respect of this conclusion. The question of

extending the industry to include pig growers, in the belief that pig growers are part of an industry producing

a “directly competitive” product, is the more contentious issue.23

The onus of proving that other products are directly competitive

Canada submits that the burden of showing products are directly competitive with other products is upon the

person asserting such a proposition. This needs to be demonstrated based upon objective evidence of the effect

of such products in the marketplace. As safeguard measures operate to relieve a country from compliance

with otherwise applicable obligations, it is appropriate that such a burden should fall on a party seeking to

widen the scope of the deviation.

The Productivity Commission has not indicated that this burden has been discharged, and the questions asked

by the Commission at the conclusion of evidence given on behalf of the Canadian Meat Council and

Canadian Pork Council on 24 August 1998 in Melbourne indicate that the matter is still at large.

21 Special Gazette No S297 of 25 June 1998 adopts the same definition, however we are unsure aboutthe implementation status of the Gazette: see Section 1.2 of this submission.22 Report No 90, January 1993.23 The Issues Paper points out that the “close processed agricultural goods” concept under Part XVBof the Customs Act 1901 does not apply to safeguard measures. In fact there is good reason to doubt theconformity of that concept with the Anti-Dumping Code in any event: see, eg, Replies to Questions Posed bythe European Communities Concerning the Notification of Australian Laws and Regulations (G/SCM/W.121,20 October 1995) at page 8, Q.B.


How “directly competitive products” should be interpreted

Can it be said that pig growers produce a product which is “directly competitive” with cuts of pork? Before

one considers that question, it must be remembered that “horizontal” competition is more readily

comprehensible as a basis for finding that products are “directly competitive” with each other. There are a

number of products which arguably are “directly competitive” with processed pork in this context. There is

ample evidence that the prices of beef and other meats affect the market demand for pork and hence pork

prices.24 Thus Canada argues that the Productivity Commission must give no less consideration to the

proposition that the domestic industry for the purposes of this inquiry should include processors of all meats

than to the more difficult proposition that it includes upstream producers of input products.

GATT practice has been to examine questions of the similarity between products on a case by case basis. This

practice under the predecessor treaty to the GATT 1994 has been followed by the WTO Appellate Body in

Japan - Taxes on Alcoholic Beverage25 (“the Japanese Liquor Tax case”).

The Japanese Liquor Tax case is instructive for the purposes of this case in two respects. Firstly, the case

continued the GATT practice of approaching issues of likeness and any extension of that term (in the

Japanese Liquor Tax case “directly competitive or substitutable”) on a case by case basis. Secondly, the

Appellate Body endorsed the approach taken in a Working Party Report on border tax adjustments that some

of the criteria useful for determining, on a case by case basis, whether a product is similar, in the sense of

being “directly competitive or substitutable” include the product’s end use; consumer’s tastes and habits; and

the product’s properties, nature and quality.26

Is industry structure relevant to determining direct competitiveness of products?

At the hearing in Melbourne on 24 August 1998, the Commission suggested that interested parties might like

to consider whether the degree of vertical integration evident in the Australian industry could lead to a

broader definition of the relevant industry. More specifically, if there were a finding of fact that pig growers

sold pig carcasses rather than live pigs, could one conclude that the pig growers themselves produce like or

directly competitive goods?

Although there is some evidence of vertical integration both upstream and downstream from pork processors,

the Australian industry is by no means totally integrated. In any event, we do not believe that vertical

integration between two consecutive but distinct stages of production is a valid reason for declaring that the

output of the second stage directly competes with the output of the first. Inputs and outputs are closely linked,

and the price of one may affect the price of the other, but they are not like or directly competitive goods as the

terms are used in the WTO simply by reason of some corporate or legal integration of productive processes by

firms in an economy.

24 See eg “Fierce competition between red and white meats”, ABARE media release dated 3 February1994; Australian Commodities, June quarter 1998 at pages 161 to 168; Pigs and Pigmeat, IndustryCommission Research Project (30 October 1995) at page 19; “Medium term policy baseline”, Agricultureand Agri-Food Canada, April 1998 pages 11 to 16.25 WT/DSC/AB/R; WT/DS10/AB/R; WT/DS11/AB/R; AB-1996-2; 4 October 1996; at page 22 et seq.

26 Tariff classifications were also considered relevant to this consideration. In this regard Canada notesthat the imported processed pork cuts under inquiry fall within tariff item 0203.29; that carcasses and halfcarcasses fall within tariff items 0203.11 or 0203.21 (depending on whether they are fresh, chilled or frozen);and that live swine fall within tariff item 0103.


This view is supported by jurisprudence in respect of “like goods” in the anti-dumping and countervailing

duty (“AD/CVD”) context. It is generally standard practice in AD/CVD cases involving vertically integrated

industries to look at the impact of imports on profits, prices, employment and other indicia on that part of the

industry’s activities that relates directly to the goods under investigation and to exclude from the analysis

other business activities in which the firms in the industry are engaged.27

It must be accepted logic that the extension of the investigation under the Agreement on Safeguards to a

consideration of “directly competitive products” is first a question of determining whether products are

directly competitive, and only secondly a question of defining the industry which produces those products.

One striking aspect of submissions presented to the Productivity Commission by Australian interests has been

the automatic or almost automatic assumption made that pig growers produce a product (dead or alive) which

is entitled to be protected from the alleged effects of imports of a downstream but very different product.

At this point it is worthwhile to remember that Canadian imports are cut into particular shapes, boneless and

frozen. The first question the Commission must ask itself is “What product or products are like or directly

competitive to these imports?” The fact that there is an alternative available does not, without more, make

that alternative directly competitive. There can be no direct competition, in the mind of an Australian buyer,

between a live pig or a carcass on the one hand, and a deboned (and, in some cases, defatted) frozen pork cut.

We remind the Commission that, as well as these differences, Canadian imports must be cooked in

accordance with Australian quarantine regulations following importation.

Based on the above, Canada submits that the fact that pig farmers and slaughterhouses may have common

ownership or that farmers may retain title to a pig until after slaughter does not transform the commercial

activity of pig farming into an industry producing a product which is directly competitive with the different

product produced by the pork processing industry. In case the Commission disagrees with this proposition,

and thinks that pigmeat in carcass form is directly competitive, then in the alternative the commercial activity

of pig slaughtering would be a possible candidate for inclusion since its output is pigmeat. However, Canada

submits that it would be incorrect to somehow transform pig growers into an industry group producing

pigmeat as a directly competitive product, as it is not that group which produces that product in the sense of

slaughtering and eviscerating it and putting it in a condition where it is directly competitive with pork cuts.

If a company participating in the production and sale of pigs is vertically integrated, then the Commission’s

injury analysis will need to carefully isolate only that part of the company constituting the industry producing

the like or directly competitive products. In the AD/CVD context, this is standard practice. In Australia, for

example, production, distribution and other functions are carefully separated to assess the impact on the

industry producing (in that context) the “like goods”. We are not aware of any case where the impact of

dumped imports of a finished product is assessed by taking into account the financial performance of

Australian suppliers of componentry to the Australian industry producing the “like” finished product. Our

point is that integration of an industry cannot broaden the definition of the industry for safeguard purposes.

The important question is whether goods are “like” or “directly competitive”, and once this is established the

industry producing those goods is the relevant industry for injury analysis.28

27 In other words, it is the business activity in relation to the production of the like goods that must beconsidered.28 In this regard, we note that the testimony on behalf of Bunge Meat Industries in Melbourne on 24August 1998 indicated that a segmented view of the company was one shared by the witness: figures were


The pork processing industry, perhaps extending to the pig slaughtering industry, is

the relevant industry

The Productivity Commission rightly points out in its “Issues Paper” (Box 1) that the term “directly

competitive” allows for a broader interpretation of an industry than does the term “like goods” used in

AD/CVD cases. Canada submits that the relevant industry is the pork processing industry, and that this could

only be extended to the pig slaughtering industry if the Commission believed that pigmeat was a directly

competitive product. Common ownership of integrated production processes is irrelevant to this conclusion,

and if there is such common ownership the Productivity Commission must separately identify that part of the

economic unit which properly falls within the relevant industry. 29

It would, in fact, make more sense to broaden the definition to include all meat processing and production

than it would to include pig growers with processors as together constituting the domestic industry. Within

the pigmeat sector, it is Canada’s considered view that the definition of the Australian industry “that

produces like or directly competitive products” that would be consistent with the Agreement on Safeguards

would be the industry that produces and sells pigmeat. This would include pigmeat destined for the fresh

market as well as that for further processing into bacon, hams and smallgoods. 30 It would be incorrect and

incongruous to extend the industry to that producing pigmeat on the basis that pigmeat was directly

competitive with imported pork, but then to try to divide the industry producing pigmeat on the basis of what

later happens to it. Extending the industry any further would involve drifting well outside the “directly

competitive products” focus of the Agreement on Safeguards.31

Canada does not agree that the Australian industry can include pig growers under the applicable Agreement

on Safeguards tests. That may be an appropriate definition for the purposes of “piggery to plate” analyses, or

research projects which are intended to consider factors impacting on a total production chain and levels

within it, but it is not an appropriate definition for the purposes of safeguards measures targeted at particular

products under the GATT 1994.

The Agreement on Safeguards test of “serious injury”

Not just increased imports, but increased imports having a particular effect

The cause of injury which must be identified to justify the application of safeguard measures under Article

XIX is “increased quantities of imports”, coupled with the superadded requirement that these increased

quantities are being imported “under such conditions” so as to have caused or threaten a particularly grave,

presented “all bar the processing segment” (Transcript, page 76); reference was made to “...our companynow at the Don Smallgoods level” (Transcript, page 83); it was concluded that “...the processors areprobably experiencing a reasonable time at this stage” (Transcript, page 76); and reference was made to “awholesaling operation in Melbourne” (Transcript, page 75).29 The Commission may also need to explore what is meant by so-called “vertical integration” .Companies or groups of companies which say they are “integrated” are nonetheless separated into corporateor divisional entities suited to the activities performed (see footnote 28).30 On this measure, imports from Canada in their peak in 1997 accounted for less than 5% ofAustralian production. Methodology: 1997 pigmeat production at 332,357 MT carcass weight, from ABS.Imports of boneless pigmeat from Canada at 9000 MT converted to carcass wt equivalent of 15,254 MT usinga factor of .59. This factor has been calculated by the Canadian Pork Council and the Canadian Meat Council,and reference should be had to their submission to the Productivity Commission for details.31 In any event, Canada requests careful verification by the Commission of testimony on 24 August1998 to the effect that “normally the first change of ownership would be at the scale, where the animal hasbeen slaughtered and eviscerated” (Bunge Meat Industries, Transcript at page 74).


or “serious” , injurious effect on the domestic industry. In other words, it is not enough to point to increased

quantities of imports: the conditions of their importation, and other conditions, must be taken into account in

determining whether there is a causal link between their importation and the alleged injury.

In this regard, we note the Commissioner’s postulation that any increase in imports might have the potential

to trigger the requirement under the Agreement of “increased quantities”.32 In isolation this may be correct,

however we believe that the sense of “emergency” from Article XIX and the strong linkage which must be

demonstrated between serious injury and imports imply that nothing short of imports with a reasonable

potential of causing serious injury qualify as the type of increase to which the Agreement might apply. In this

context normal or predictable variances could not be expected to trigger safeguard measures. Article 4.2(a)

speaks of the need to evaluate “the rate and amount” of the increase in imports “in absolute or relative

terms”, and the “share of the domestic market taken by imports”. We suggest that there would be no need to

evaluate these factors if any increase, no matter how small, could trigger safeguards measures. Even if this is

not clear to the Commission, the need to attribute serious injury to increased imports expresses the need for

there to have been an increase of sufficient magnitude to cause such an effect. Essentially, it is the measuring

of this effect which calls for an appreciation of the magnitude of the increase.

The Commission’s suggestion that the Agreement on Safeguards does not specify how much of an increase in

imports is required in order for the Agreement to apply was made in isolation to the questions of serious

injury and the causal link which needs to be demonstrated between increased imports and that injury. The

Commission accepted that an increase in imports was required but questioned whether this necessarily needed

to be an increase of any particular quantum. The Commission noted that it did not see that any particular

quantum was specified.

Canada concurs that the Agreement does not address the issue of the amount of increases in imports as a

threshold question other than the fact that there must be an increase. The test in the Agreement is not the

amount of increased imports but the effect that occurs as a result of increases in imports. The test as set out in

Article 2.1 of the Agreement is that "product is being imported... in such quantities... as to cause or threaten

to cause serious injury to the domestic industry...".

Canada notes that the issue to be addressed is whether an increase in imports has occurred and whether there

is a causal link to serious injury or threat thereof to the industry. However, Canada would suggest that the

quantum of increased imports is an important fact to be determined. While each case needs to be considered

on a case by case basis and with due regard to the facts in each case, it would be very unlikely that a minimal

increase in imports would cause serious injury. Such a minimal increase would also not appear to be a

situation which could be characterised as leading to an "emergency" or as an “unforeseen development”.33

The actual conditions of the importation of the like products, and other conditions affecting the domestic

industry, must be considered in their totality. We reconsider this proposition when dealing with the

methodology for assessing the other competitive factors and causes of injury which impact upon the

Australian industry. 34

32 Transcript of Proceedings, Melbourne 24 August 1998 at page 114.33 See 6.4 below.34 See 4.3 below.


The seriousness of “serious injury”

The Issues Paper correctly points out that the Agreement on Safeguards imposes a more stringent injury test

than that which applies in the case of investigations as to whether injury has been caused by dumped or

subsidised imports. Whilst little guidance is provided in quantitative terms within the text of the Agreement

on Safeguards itself, there are a number of indications in the Agreement as to the way in which serious injury

must be considered, and there are a number of aids to interpretation which can help with the task of defining

an appropriate threshold. 35

Under the Agreement, it is said that “serious injury shall be understood to mean a significant overall

impairment in the position of a domestic industry”. The use of these words “a significant overall

impairment” are instructive. They emphasise the pervasive impact that imports must be found to have had on

the entirety of the domestic industry before safeguard measures might be applied. This test is to be contrasted

with the lesser test of material injury in the dumping context.

Article 4.1(c) of the Agreement on Safeguards emphasises that the domestic industry, which must be

significantly impaired in an overall sense, is all of the producers (ie “the producers as a whole”) of the like or

directly competitive products, or those whose collective output constitutes a major proportion of the total

domestic production of those products.

Causal link, and non-attribution of other factors

Serious injury must be caused by increased imports

Article 4.2(a) also emphasises how strict the causal link test is under the Agreement on Safeguards, requiring

the competent authorities to “evaluate all relevant factors of an objective and quantifiable nature”. Article

4.2(b) says that the causal link must be demonstrated on the basis of objective evidence, and that if factors

other than increased imports are causing injury to the domestic industry, such injuries shall not be attributed

to increased imports.

Article 2 of the Agreement sets out a number of conditions that must exist for a member to apply safeguard

measures. The structure of that Article is extremely significant as is the fact that the Article is entitled

"Conditions". The title is part of the text and thus meaning must be ascribed to it. One of the common and

ordinary meanings of the word condition is "prerequisite". That meaning is reinforced by the language of

paragraph 2.1. This permits a WTO Member to apply safeguard measures "only if" the Member has

determined the existence of certain factual matters. The three set out in that paragraph are:

• increased imports;

• under such conditions as to cause or threaten to cause; and

• serious injury to the domestic industry that produces like or directly competitive

products.

The rules for the determination of serious injury are found in Article 4 of the Agreement. Article 4.2(b)

further reinforces the necessity for there to be a causal link between increased imports and serious injury.

35 In Stewart’s The GATT Uruguay Round, A Negotiating History (1986-1992) Vol. II: Commentary ,the author draws attention to the fact that of the 62 safeguard (“escape clause”) actions up to 1990 in theUSA, there were affirmative US International Trade Commission decisions in fewer than half, andPresidential relief was ordered in only 11.


That subparagraph specifically directs that a determination of serious injury shall not be made unless “the

existence of a causal link” can be demonstrated "on the basis of objective evidence”.

The effect of other injury factors must not be attributed to increased imports

Subparagraph 4.2(b) gives further guidance to domestic investigating authorities with respect to other factors

for which there is no causal link between increased imports and serious injury or threat thereof. The second

sentence of the subparagraph reads as follows:

"When factors other than increased imports are causing injury to the domestic

industry at the same time, such injury shall not be attributed to increased imports."

In order to establish a causal link it is not sufficient for a domestic investigating authority to only consider

whether the industry concerned is suffering injury. Rather, there needs to be an analysis of the factors which

are causing the alleged injury. The Agreement recognizes that other factors could be causing, or at least

contributing to injury being felt by the domestic industry. The Agreement requires that injury caused by

factors other than increased imports is to be disregarded for purposes of attributing what is the injury which is

attributed to increased imports.

What is it that is not to be attributed to increased imports? It is submitted that the second sentence of Article

4.2(b) is grammatically clear. The sentence begins with a discussion of factors other than increased imports

that are causing injury. The sentence then states that "such" injury cannot be attributed to increased imports.

To what does "such" refer? The Shorter Oxford Dictionary36 gives this description:

"Standing predicatively at the head of a sentence or clause, and referring

summarily to a statement or description just made;.Of the same kind or class as

something mentioned or referred to;.Equivalent to a descriptive adj. or adv. on

which it follows closely and the repetition of which is thus avoided.”

Black’s Law Dictionary37 notes:

"‘Such’ represents the object as already particularized in terms which are not

mentioned and is a descriptive and relative word, referring to the last antecedent."

In the context of the second sentence of subparagraph 4.2(b) it is clear that "such" injury in the concluding

clause of the sentence is a grammatical link to the previous reference to injury. That reference occurs in the

first portion of the sentence which is addressing injury caused by factors other than increased imports. In the

result, injury that is caused by factors other than increased imports cannot be attributed to increased imports

(ie cannot be part of the injury which is caused by increased imports).

The proper methodology for determining whether a causal link exists

Canada submits that the practical result is that a domestic investigating authority needs to analyse, in terms of

causation, the injury from which the domestic industry is said to be suffering. If injury is being caused by

factors other than increased imports then that injury cannot be attributed to imports. Its only relevance to the

question of demonstrating injury caused by increased imports is that it contradicts the proposition that

increased imports have themselves caused serious injury, and dilutes any injury caused by “increased

36 “The Shorter Oxford English Dictionary on Historical Principles” (3rd edition) prepared by C.T.Onions; Clarendon Press, Oxford, reprinted 1967.37 “Black’s Law Dictionary” (6th edition) Henry Campbell Black and Sixth Edition by Editorial Staff,West Public Company, St. Paul, Minnesota, USA; 1990.


imports”. Similarly, if injury exists but it is unclear as to whether or not it has been caused by increased

imports, Canada submits that such injury must be regarded as injury which cannot be attributed to imports. In

the latter case it could not be said that the investigation has met the burden of "demonstrating", on the basis

of objective evidence, the existence of a causal link.

It may be that increased imports are one factor among many that is causing (or threatening to cause) injury to

the domestic industry. If the competent authorities reach such a conclusion, they must then consider whether

that injury and only that injury (ie, injury attributable to increased imports) meets the test of "serious injury"

which is set out in Article 4.1(a) of the Agreement. Injury attributed to other factors cannot be part of "serious

injury." Having identified an injury attributable to increased imports, the next step that the competent

authorities need to do is analyze whether the injury amounts to "serious injury". That is, is the injury that has

been determined to be attributable to an increase in imports such as to amount to a significant overall

impairment in the position of the domestic industry?

During the Public Hearing at Melbourne the Commission gave some consideration to the methodology for

determining whether there is a causal link between increased imports and serious injury. The Commissioner

said that it must be established “that any injury, the serious injury, that the industry is trying to establish..

must be serious injury attributable to an increase in imports”. The Commissioner also ventured to suggest

“[t]hat doesn’t mean that there may not be other causes of serious injury but there has to be enough, if you

like, serious injury attributable to the increased imports in order to sustain the argument.” These

observations may be technically correct, however Canada wishes to make a number of points in relation to

them.

First, the Agreement on Safeguards requires a “significant overall impairment in the position of the domestic

industry”. This would appear to suggest that an injury finding could not be made where injury was suffered

by some parts of an industry,38 but not other parts.

Secondly, it would be difficult to come to a finding that there were a number of competing causes of injury,

each of which caused serious injury. If three or four separately identifiable “injury” effects could be

identified, the fact that there were so many would seem to mitigate any reasonable conclusion that each injury

effect was serious. Otherwise, the extreme nature of the “serious injury” requirement would be diluted.

Furthermore, in the case of a number of factors all causing “serious injury” , the structural adjustment

contemplated by Article 5.1 of the Agreement on Safeguards could not be achieved in the future because

safeguard measures would not prevent or remedy serious injury.

38 Or, at least, that part o f an industry falling short of the “major proportion” referred to in Article4.1(c).


In the dumping context, the principal authority on the question of the necessary causal link between dumped

imports and material injury is ICI Australia Operations Pty Ltd v Fraser and Ors.39 The salient quote from

the Full Court’s decision is the following:

“Where the Australian industry under consideration has suffered detriment from a

number of causes, it will be necessary for the Minister to be satisfied that the

industry has suffered detriment sufficient to meet the description “material injury”

within the meaning of the legislation in consequence of the dumping of goods that

have been exported to Australia, and to quantitatively separate that material injury

from detriment caused by other factors. Whether the Minister is so satisfied as to

the first matter involves a matter of judgement and degree, and is a question of

fact... Understood in this way, it is correct to say... that [the relevant section]

requires that the “material injury” referred to must be caused solely by the

dumping the subject of the inquiry”.

Canada believes this is a legally correct methodology to adopt in deciding whether the causal link exists

between serious injury and increased imports. However, it is even more stringent in the safeguard context

because of the requirement that the injury entail “a significant overall impairment in the position of a

domestic industry”. A finding of a causal link between increased imports and material injury in the anti-

dumping context may be open where the material injury has occurred to one sector of the industry. 40

However, “significant overall impairment” adds a different dimension to the severity of the injury concerned.

Threat

Article 4.1(b) requires any threat of serious injury to be “clearly imminent”. The determination of this threat

must be “based on facts and not merely on allegation, conjecture or remote possibility.”

No less rigour must be devoted towards the analysis of injury factors and their impact in respect of what

might happen in the future than the rigour which is required in respect of a present “serious injury”

determination. This is clear from the linkage between Article 4.1(b) and Article 4.2.

The inter-relationship between Article XIX and the Agreement on Safeguards

On 24 August 1998 during the public inquiry in Melbourne in this matter, the Commission queried the

relationship, if any, between Article XIX of the GATT 1994 and the Agreement on Safeguards which

elaborates upon the substantive tests and procedural requirements of a safeguards investigation.

The GATT 1994 is a carefully integrated trade agreement

By virtue of Article II:2 of the GATT 1994, the WTO Agreement together with the Multilateral Trade

Agreements annexed thereto form an integral package. 41 In Brazil - Measures Affecting Desiccated

Coconuts42 ("the Coconuts case"), the WTO Appellate Body stated as follows:

39 (1992) 106 ALR 257.40 Swan Portland Cement Limited & Anor v The Minister for Small Business and Customs & Anor(1991) 28 FCR 135.41 Article II:2 of the WTO Agreement reads as follows:

“The agreements and associated legal instruments included in Annexes 1, 2 and 3(hereinafter referred to as “Multilateral Trade Agreements”) are integral parts of thisAgreement, binding on all Members.”

42 AB-1996-4; WT/DS22/AB/R; 21 February 1997.


"The WTO Agreement is fundamentally different from the GATT system which

preceded it. The previous system was made up of several agreements,

understandings and legal instruments, the most


significant of which were the GATT 1947 and the nine Tokyo Round Agreement...

Unlike the previous GATT system, the WTO Agreement is a single treaty

instrument which was accepted by the WTO Members as a ‘single undertaking’."

The Appellate Body went on to add that:

"The relationship between the GATT 1994 and the other goods agreements in

Annex 1A is complex and must be examined on a case -by-case basis. Although the

provisions of the GATT 1947 were incorporated into, and became part of the GATT

1994, they are not the sum total of the rights and obligations of WTO Members

concerning a particular matter... The general interpretative note to Annex 1A was

added to reflect that the other goods agreements in Annex 1A, in many ways,

represent a substantial elaboration of the provisions of the GATT 1994."

As such, Article XIX of the GATT 1994 cannot be viewed as being separate and distinct from the Agreement

on Safeguards. Rather, the Article and the Agreement form an inseparable package of rights and obligations

with the Agreement representing an elaboration of the provisions of the former. 43

Meaning and effect must be given to related provisions of the GATT 1994

Article 31:1 of the Vienna Convention on the Law of Treaties provides the following general rule of

interpretation:

“A treaty shall be interpreted in good faith in accordance with the ordinary

meaning to be given the terms of the treaty in their context and in the light of its

object and purpose.”

In United States - Standards for Reformulated and Conventional Gasoline,44 the Appellate Body explained

that:

"One of the corollaries to the ‘general rule of interpretation’ in the Vienna

Convention on the Law of Treaties is that interpretation must give meaning and

effect to all the terms of a treaty. An interpreter is not free to adopt a reading that

would result in reducing whole clauses or paragraphs of a treaty to redundancy or

inutility."

Therefore, in reading Article XIX of GATT 1994 in conjunction with the Agreement on Safeguards, one

must try to ascribe a reasonable interpretation that gives meaning and effect to all the terms in both.

In the event of an unavoidable conflict, the general interpretative note to Annex 1A becomes relevant. That

note reads as follows:

“In the event of a conflict between the provisions of the General Agreement on

Tariffs and Trade 1994 and a provision of another agreement in Annex 1A to the

Agreement Establishing the World Trade Organisation (referred to in the

43 The second recital in the preamble to the Agreement on Safeguards reads as follows:

“Recognising the need to clarify and reinforce the disciplines of GATT 1994, andspecifically those of its Article XIX (Emergency Action on Imports of Particular Products), to re-establish multilateral control over safeguards and eliminate measures that escape such control;”

44 WT/DS2/AB/R, 29 April 1996.


Agreements in Annex 1A as the "WTO Agreement") the provisions of the other

agreement shall prevail to the extent of the conflict."

The Agreement builds on Article XIX, and does not replace it

Canada submits that there is no reason to doubt that the Agreement is an elaboration and an amplification of

Article XIX. Both the “emergency” character of Article XIX, and "unforeseen developments" circumstance

contained in Article XIX:1(a) of the GATT 1994 are part and parcel of the Agreement. An interpretation

which gives effect to all the terms in both instruments is clearly achieved by the linkage between Articles 1

and 2.1. Article 2.1 of the Agreement on Safeguards must be read in conjunction with the general provision

set out in Article 1 of the Agreement, which states that:

“This Agreement establishes rules for the application of safeguard measures which

shall be understood to mean those measures provided for in Article XIX of GATT

1994."

In this regard, it can be seen that the measures provided for in Article XIX are, by necessity, measures in

response to unforeseen circumstances, thereby avoiding any inconsistency between Article XIX and the

Agreement on Safeguards.

In summary, it is Canada’s view that the Agreement on Safeguards builds on but does not replace Article

XIX, and that therefore the concepts embodied in Article XIX are equally embodied in the application of the

Agreement on Safeguards.

“Emergency action” due to “unforeseen developments” is part of the Agreement on

Safeguards

In respect of the “emergency” character to the action required, we see nothing in the Agreement on

Safeguards to suggest that it is no longer directed at “emergency action”: to the contrary, there are clear

indications that Article XIX’s objectives are unchanged under the Agreement. The Recitals and Article 1

underpin the primacy of Article XIX. The second Recital refers to Article XIX, and sets out its full title of

“Emergency Action on Imports of Particular Products”. The same Recital refers to the clarification and

reinforcement of the disciplines of Article XIX, which implies that it is the emergency action which is to be

clarified and reinforced, not that the emergency requirement is able to be dispensed with. Article 1 of the

Agreement makes it clear that the Agreement amplifies, but does not derogate from, “those measures

provided for in Article XIX”.

The only applicable GATT precedent on the matter of "unforeseen developments" is a Working Party report

on Withdrawal by the United States of a Tariff Concession under Article XIX.45 The Working Party, except

for the United States, stated that:

"...the term 'unforeseen developments' should be interpreted to mean developments

occurring after the negotiation of the relevant tariff concession which it would not

be reasonable to expect that the negotiators of the country making the concession

could and should have foreseen at the time when the concession was negotiated."

45 GATT/CP/106, adopted on 22 October 1951.


“Industry” clarifies ambiguities inherent in Article XIX’s reference to “producers”

During the August 24 hearing the Commission queried why Article XIX of the GATT 1994 used the term

"domestic producers" while the Agreement on Safeguards used the term "domestic industry."

As the WTO Appellate Body noted in the Coconuts case, the WTO Agreement is a "single undertaking" and

"[w]ithin this framework, all WTO Members are bound by all the rights and obligations in the WTO

Agreement and its Annexes 1, 2 and 3" . As noted previously, the interpretation of this bundle of rights and

obligations is to be done following the customary rules of international law and in particular, the Vienna

Convention on the Law of Treaties. Paragraph 1 of Article 31 of the Vienna Convention on the Law of

Treaties establishes the primary principle that treaties are to be "...interpreted in good faith in accordance

with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object

and purpose."

In the Coconuts case the Appellate Body, as noted earlier, stated that, "[t]he relationship between the GATT

1994 and the other goods agreements in Annex 1A is complex and must be examined on a case-by-case

basis". Accordingly, one must interpret these two provisions together in accordance with their ordinary

meaning in the context of the treaty and in light of its object and purpose.

Canada submits that a key contextual element is the preamble to the Agreement which noted that WTO

Members recognised the need to "clarify" GATT Article XIX disciplines. It is Canada’s submission that the

two terms should be read in the light of the Agreement clarifying the application of safeguards.

If one were only to read Article XIX of the GATT 1994 one would conclude that one of the applicable tests

was that there needed to be a serious injury (or threat thereof) to domestic producers. However, there could be

several interpretations of what test is to be applied. Do you merely need to show, for example, that there is a

serious injury suffered by more than one domestic producer? On the other hand, is the remedy intended to be

available only if all domestic producers as a class are suffering serious injury or are faced with the threat

thereof? What if two producers are affected but they are statistically insignificant? Does the injury to these

producers permit the application of safeguard measures? What if only 2% of domestic producers are affected

but these producers produce 80% of the product in question?

The change brought about by the reference to “industry” rather than “producers” is meant to emphasise the

requirement that injury be shown to the producers “as a whole”, as is stated under Article 4.1(c) of the

Agreement. This terminology was borrowed from the Anti-Dumping Agreement to guard against complaints

being assessed with reference to producers whose output was less than a major proportion of domestic

production. “Producers” is amenable to a more selective group than the singular and collective expression

“industry” . Together with the “significant overall impairment” requirement the modified terminology

clarifies the high standard of injury to prevent abuse.

The Agreement clarifies this point by dealing with producers within the context of the "domestic industry" as

set out in subparagraph 4.1(c) of the Agreement. By reading together the applicable provisions it is clear that

the producers that are relevant are those whose output of like or directly competitive products constitute a

major proportion of the domestic production of those products. The terms are not conflicting but rather

complementary. Given the view of the Appellate Body that the WTO is a single undertaking, the terms should

be read together if it is reasonable to do so and if to do so does not offend the normal rules of interpretation.

Such an interpretation is also reasonable given that safeguards allows a Member to avoid otherwise applicable


obligations and this should only be permitted if the alleged injury is substantial. With respect to the

interpretation of these terms, Canada submits that reading these provisions together is consistent with the

object and purpose of the GATT 1994 and the Agreement as it clarifies the application of both of them.

Anti-dumping and countervailing

The Terms of Reference do not embrace the question of whether imports are subsidised or dumped and

consequently the Commission “will not be considering whether anti-dumping action may be warranted”.

There are however some observations from this field which may be of assistance in the context of its inquiry.

The Australian Customs Service found in November 1992, and the Anti-Dumping Authority confirmed in

January 1993, that:

• there was virtually no incidence or margin of dumping;

• there was no evidence that subsidies affected exports;

• Canadian pork imports caused no material injury; and

• the same imports posed no threat of material injury.

The standard of proof of injury in a safeguard action is higher than the material injury requirements in an

AD/CVD case.

Prior to and since that finding, Canadian hog producers have been operating under the close scrutiny of US

trade officials who annually review the level of alleged production assistance provided to the Canadian

industry. While the US has applied countervailing duties on live hog exports from Canada the rates in recent

years have been very small. The three recent US administrative reviews have resulted in de minimis (zero)

deposit rates being applied to live hogs imported from Canada over the period 1 April 1994 to 30 March

1997. Consequently no cash deposits on future Canadian imports will be required.

Rebuttal of economic impact study

On 19 August 1998 in Brisbane, the Commission heard testimony from Mr T Purcell of the University of

Queensland concerning a report prepared for the Queensland Department of Primary Industries entitled “The

Impact of Trade Liberalisation and Increasing Imports on Australian Pig Prices”.

Canada rejects the conclusions of that report. Attachment 1 is an independent analysis of the Purcell report

commissioned by Agriculture and Agrifood Canada, a department of the Canadian government.

Concluding comments

The Canadian pig growing and pork processing industries will be presenting information to the Productivity

Commission about the position of the Australian industry and the domestic and international commercial

environment within which it operates. On the basis of these submissions and that industry information,

Canada is firmly of the view that there can be no grounds for the imposition of safeguard measures.

In particular, but without derogating from any other relevant matters, Canada believes:

• the Australian industry does not include the activity of pig producing prior to the

processing stage of pigmeat production;

• the Productivity Commission has before it evidence of injury causing factors which

together suggest that the small level of Canadian imports cannot be a cause of

serious injury; and


• the small level of Canadian imports cannot be considered to be a development

which is in any way unforeseen following the relaxation of quarantine controls or

imports of Canadian pork in June 1990.

Accordingly, Canada respectfully requests the Commission to report to the Treasurer that circumstances do

not justify safeguard measures to be imposed on imports of meat of swine, frozen, falling within tariff sub-

heading 0203.29 of the Australian Customs Tariff under the WTO Agreement as referred to in Gazette S297

of 25 June 1998. It is unnecessary for the Productivity Commission to consider measures which would be

necessary to prevent or remedy serious injury and to facilitate adjustment.

Submitted for and on behalf of theGOVERNMENT OF CANADAby its Australian Counsel

Daniel MoulisPartnerFREEHILL HOLLINGDALE & PAGE

18 September 1998


Table of contentsPage

1 Introduction 1

1.1 Canadian pork exports to Australia face yet another inquiry 11.2 Proper procedures must be followed by the Productivity Commission 21.3 The keypoints of this submission 21.4 Agreement on Safeguards issues addressed by this submission 3

2 What is the domestic industry producing like or directly competitiveproducts? 4

2.1 The Terms of Reference refer to a specific category of goods 42.2 What are the “like goods” to imported boneless frozen pork? 52.3 The onus of proving that other products are directly competitive 52.4 How “directly competitive products” should be interpreted 62.5 Is industry structure relevant to determining direct competitiveness ofproducts? 62.6 The pork processing industry, perhaps extending to the pig slaughteringindustry, is the relevant industry 8

3 The Agreement on Safeguards test of “serious injury” 8

3.1 Not just increased imports, but increased imports having a particular effect 83.2 The seriousness of “serious injury” 10

4 Causal link, and non-attribution of other factors 10

4.1 Serious injury must be caused by increased imports 104.2 The effect of other injury factors must not be attributed to increasedimports 114.3 The proper methodology for determining whether a causal link exists 11

5 Threat 13

6 The inter-relationship between Article XIX and the Agreement on Safeguards 13

6.1 The GATT 1994 is a carefully integrated trade agreement 136.2 Meaning and effect must be given to related provisions of the GATT 1994 156.3 The Agreement builds on Article XIX, and does not replace it 166.4 “Emergency action” due to “unforeseen developments” is part of theAgreement on Safeguards 166.5 “Industry” clarifies ambiguities inherent in Article XIX’s reference to“producers” 17

7 Anti-dumping and countervailing 18

8 Rebuttal of economic impact study 18

9 Concluding comments 18


In the Productivity Commission

Inquiry into Pig and Pigmeat IndustriesSafeguard Action Against Imports

Submission of the Government of Canada

London Court 13 London Circuit Canberra City Australian Capital Territory 2601 AustraliaTelephone (02) 6240 6100 Int + (61 2) 6240 6100 Facsimile (02) 6240 6222 DX 5666 Canberra

Reference: Daniel Moulis

SY D N E Y ME LB O U R N E P E R TH C AN B E R R A B R ISB AN E SIN G AP O R E H AN O I H O C H I MIN H C ITYC O R R E SP O N D E N T O F F IC E IN J AK AR TA

Liab ilit y is lim it ed b y th e So lic it o rs Sch em e u n d er t h e P ro fes s io n a l S t an d ard s Ac t 1994 (NSW )

1


ATTACHMENT 1

An Assessment of Tim Purcell and Steve Harrison’ report entitled:“The Impact of Trade Liberalization and Increasing Imports on

Australian Pig Prices”

by

Bruno LarueAssociate Professor

Centre de Recherche en Economie Agroalimentaire (CREA)Université Laval, Ste-Foy, Quebec

Canada, G1K 7P4 andDepartment of Economics

Iowa State University, Ames, IowaUSA, 50011

September 10, 1998

Report prepared for the International Trade Policy Directorate ofthe Market and Industry Services Branch of Agriculture and Agri-

Food Canada

Acknowledgements: This assessment was prepared while I was onsabbatical leave. I would like to thank the department ofEconomics at Iowa State University for the resources madeavailable to me. I also wish to acknowledge the usefuldiscussions I had with Professors Harvey Lapan and Marvin Hayengaregarding safeguard issues, model specification and the hog/porkindustry in Australia. The opinions expressed in my assessmentare mine and need not be shared by Professors Lapan and Hayenga.

2


EXECUTIVE SUMMARY

In order to justify a safeguard action , it is necessary to showthat imports of a product are increasing and that domesticproducers of competitive products are seriously injured orthreatened with serious injury, and that this injury or threat iscaused by the increased in imports. As implied by the title oftheir report (i.e., “The Impact of Trade Liberalization andIncreasing Imports on Australian Pig Prices”), Purcell andHarrision (P&H) produced a study narrowly focused on prices.Having gone in that direction, they cannot provide estimatesabout the degree of injury suffered. Their analysis should havebeen extended to measure the welfare effects of pork imports onAustralia’s hog/pork industry. This would have been a much moreambitious endeavor because it requires detailed modeling of allthe components of the industry (retail, pork processing, hogproduction and feed supply). The authors failed to analyze theconsequences of changes in imports, hog prices and production andtherefore did not prove that serious injury was inflicted nor didthey provide sufficient empirical evidence to guide policymakers.Economists have developed welfare measures that are well-suitedto assess the extent of the injuries or benefits stemming fromimports. These measures can be used to compute the welfarechanges all along the marketing chain, from the consumers of meatproducts all the way down to hog producers and their inputsuppliers. By not providing the necessary elements to assess thedegree of injury suffered by the Australian industry, the P&Hreport is of little utility to evaluate the merit of a safeguardaction. 46

The P&H report is long on econometric procedures but short oneconomic analysis. Nevertheless there are problems with thechoice and implementation of econometric techniques but thesetend to be minor in comparison to the model specificationproblems encountered. The lack of a solid theoretical foundationthat led to the exclusion of relevant variables and the inclusionof irrelevant ones, make the interpretation of the models verydifficult, the link with economic theory being lost. Also, theexclusion of relevant variables causes the estimated coefficientsin a regression to be biased. P&H relied on single equationestimation techniques when there were obvious endogenousvariables on the right hand side. As a result, their estimatedcoefficients suffer from an endogeneity/simultaneity bias thatfurther erodes the confidence one might have in their results.The inappropriate model specifications and estimation problems ofP&H produced unreliable results.

The results from the VAR modeling (p.76-77) and those of thesingle equation supply model (p.55-56) produced mixed resultsregarding the significance of the effect of Canadian imports onAustralian economic variables (like the price of pork).However, both types of models suggest that Canadian and

46 Hoeckman and Kostecki (1995, p.193) argue that the injury criteria should besuch as to ensure that the domestic industry as a whole is making substantiallosses.

3


Australian pork prices are independent. The P&H interpretationof unit root tests is incorrect because testing whetherinnovations have a permanent effect on the level of a series,however interesting, is simply impossible to do. Finally, thediscussion on the economics of safeguards and tradeliberalization is flawed and incomplete. Overall the evidencepresented in the P&H report is weak and unconvincing.

1. INTRODUCTION

The report prepared by Tim Purcell and Steve Harrison (P&Hhenceforth) was requested from the Queensland Department ofPrimary Industries, to investigate the impact of changes inregulations for pigmeat imports and increased import quantitieson the Australian pigmeat industry. The report attempted toanswer four questions (p.3): What has been the underlying patternof change in Australian pigmeat prices? 2) Have pigmeat importsled to lower domestic prices? 3) How adequate have themethodology and conclusions drawn in previous studies of theimpact of imports on the Australian pig industry been? Shouldthe Government wish to provide assistance to the industry in thecurrent state of extremely depressed prices, what considerationsneed to be taken into account?

The purpose of the P&H report is to provide documented empiricalevidence to justify a safeguard action, a set of governmentactions responding to imports that are deemed to harm theimporting country’s economy or domestic competing industries.The safeguard action would allow Australia to use means tocurtail imports of pork to protect its allegedly injured domesticindustry for as long as eight years. 47 Canada is targeted by suchaction because it is the main supplier of pork imports inAustralia.

According to the WTO agreement on safeguards, a Member may applya safeguard measure to a product only if that Member hasdetermined, pursuant to the provisions set out below, that suchproduct is being imported into its territory in such increasedquantities, absolute or relative to domestic production, andunder such conditions as to cause or threaten to cause seriousinjury to the domestic industry that produces like or directlycompetitive products. Thus a country wanting to use Article XIXmust show: 1) that imports have increased; and 2) that domesticproducers of competitive products are seriously injured orthreatened with serious injury, and that this injury or threat iscaused by the increased in imports. Jackson (1997) argue thatthe application of this two-step rule is problematic because someof the key elements are ambiguously defined (i.e., industry, likeand competitive products, injury, threat…). Sampson (1987) andFinger (1996) are less generous and contend that historicinterpretations of these two conditions have made it possible forany increase in imports to be actionable under Article XIX prior

47 The 1994 WTO agreement on safeguards stipulates that an importing country mustprogressively reduce the restrictiveness of its safeguard measures. The durationof safeguard measures is limited to four years with one possible extension.

4


to 1994. This loophole seems to have been closed in the wake ofthe 1994 WTO Agreement on Safeguards. It is important to notethat in a safeguard case there is no need to establish theexistence of unfair trade practices such as export subsidies ordumping as imports need not be ‘unfair’ to cause injury.Accordingly, lobby groups seeking protection might requestsafeguard actions when proofs of improper trade practices are notsolid enough to support countervailing and antidumping actions.Indeed, such avenues had been unsuccessfully pursued by theAustralian pig industry. However, not having to document unfairtrade practices does not mean that protectionist interests caneasily secure import protection through the escape/safeguardclause. One might hope that the escape-clause standard forinjury should be the highest or most difficult to establish,given that the escape clause is designed to respond to situationsthat do not necessarily involve any unfair action by foreignexporters. However, it is the issue of compensation orreciprocity that has apparently restrained the number of ArticleXIX actions in the past (Finger, 1996, p.321). On that score,the new limits imposed on retaliation by the WTO agreement onsafeguards (i.e., an exporter is prevented from retaliating forthe first three years that a safeguard measure is in effect) mayfacilitate government capture by protectionist lobbies. It ishoped that improvements in procedural requirements embodied inthe WTO agreement of safeguards will prevent abuses of ArticleXIX.

It is critical for the Australian pig industry to demonstratethat their industry was seriously injured and that imports werethe main culprit if their quest for a safeguard action is to besuccessful. The P&H report brings empirical evidence ofquestionable quality to document that imports have had a negativeimpact on hog prices and production in Australia. In whatfollows, I will provide a detailed assessment of the theoreticaland empirical evidence presented in the P&H report. In the nextsection, I address the theoretical effects of tradeliberalization, adjustment policies and protection. The issuesraised are critical to all countries concerned. My review of theliterature of the economics of safeguards reveals that safeguardmeasures are rarely motivated by sound economic principles.Section 3 describes the content of the P&H report and themethodology chosen by the authors. It is argued that theirframework cannot provide the necessary elements to establish thatimports of pork have caused serious injury to the Australianindustry. Section 4 critically reviews what P&H tried to measureand uncovered model specification and econometric problems.Section 5 concludes and summarizes my main objections to the P&Hreport.

2. TRADE LIBERALIZATION, ADJUSTMENT POLICIES AND PROTECTION

The optimality of a free trade policy for a small open economy isa proposition that makes unanimity among international tradeexperts. It is discussed in virtually all textbooks (e.g., Dixitand Norman, 1980) and it can be traced back to an earlycontribution made by Paul Samuelson (1939). In the absence of

5


externalities and market power, production and consumptiondecisions are most efficient when based on exogenously determinedworld prices. In other words, free trade maximizes the gainsfrom trade. Of course, there are “accepted” exceptions to thisgeneral principle. Choi and Lapan (1991) have shown that when apublic good can only be financed through commercial policy, asmall tariff is likely to be optimal. This case does not applyto a developed economy like Australia’s. Johnson (1951) and afew others before him have shown that when a country caninfluence its terms of trade, an optimal tariff can improvewelfare. The so-called optimal tariff argument does not applyhere for two reasons: Australia is not likely to be able toinfluence the international price of pork products and a tariffwould most likely trigger (after three years) reprisals whichwould adversely affect its welfare. Corden (1984) in his reviewof the normative theory of international trade points out thattariffs and quotas are rarely first best policy instruments toaddress domestic market disruptions or to achieve non-economicobjectives like sector specific production or employment targets.However, trade barriers are first best instruments to achieve thenon-economic objective of import reduction (Corden, 1984;Vousden, 1990). In cases like that, a government is willing totrade off economic welfare to achieve a political target, hencethe term non-economic objective.

Some economists have also analyzed specifically the economics ofsafeguards. For most, safeguard measures are difficult tojustify on economic grounds.

Trade policy is usually inefficient in that it tends tocreate more distortions than it solves. Indeed, Deardorffand Stern have likened trade policy to doing acupuncturewith a two-pronged fork; even if one of the prongs finds theright spot, the other prong can only do harm. This appliesto protection to market disruption as well.

Hoeckman and Kostecki (1995, p. 194)

GATT safeguards make no economic sense… In economics,safeguards are nothing but ordinary protection withsufficient political muscle to prevent antiprotection forcesfrom interfering.

Finger (1996, p.334)

There are essentially two approaches to the analysis of safeguardmeasures. There is a normative approach strictly concerned withefficiency and national welfare and a positive approach based onCorden’s social welfare function which posits that a governmentis willing to trade off efficiency for equity. Economists havebeen concerned for some time about adjustment problems as theeconomy moves from one equilibrium to another. Usually,adjustment problems are linked to factor specificity or limitedmobility, factor price stickiness, or congestion in job search.In cases like that, the first best policies are not tradepolicies. Mayer (1994, p.315) summarizes well the literature

6


when he states that "the efficiency argument (for safeguard) isquite weak, however, as it is valid under rather specialcircumstances only and no intervention is the general rule." Heconstructed a model (in which intervention can increase welfare)to compare tariffs, income transfers to workers of the injuredindustry and income transfers to people who opted for retraining.His ranking classifies the tariff as the third best instrument.Mussa (1982) on the other is less enthusiastic about the use ofsafeguard measures to ease adjustments. His objections arecompletely efficiency-related. He showed that private maximizingbehavior will lead to socially efficient adjustement processprovided that the prices of outputs and factors and the discountrate perceived by private agents correspond to their true socialvalues, and provided that the expectations that influence privatedecisions are rational. In contrast, Deardorff (1993) espousesa purely normative/political economic approach by assuming thatthe motivation behind a safeguard action is to avoid (and notease or facilitate) adjustments. The rationale is that increasesin imports may be temporary and need not reflect permanent shiftsin comparative advantage. He and Hoekman and Leidy (1993) favorthe use of transferable export licenses to curb foreign export inan efficient and equitable manner. Deardorff’s argument issimilar to the well known “tariff as insurance” argument whichhas lost much support among economists (see Vousden, 1990).Furthermore, the ubiquity of revenue or income insurance programsto help agricultural producers cope with risk makes Deardorff’sargument irrelevant. Revenue or income insurance programs forsure dominate tariffs or quotas in this case.

The P&H report does not dwell very much on trade theory and whenit does it arrives at bizarre conclusions. For example on p.104, the analysis of the gains from trade associated with thedeparture from autarky to free trade is plain wrong. In thispartial equilibrium framework, the gains are to be evaluated interms of changes in consumer and producer surpluses. Thevariation in consumer surplus is the area under the demand curvebetween the autarky (closed economy) price and the free tradeprice, that is area Pe, Pw, g, f. The loss in producer surplusis the area above the supply curve between the autarky and freetrade prices, that is area Pe, Pw, e, f. The net gain (fromtrade) is positive and is defined by area f, g, e. Thesubsequent argument for subsidies (see 6.3.2) is also poorlyjustified. The argument made by P&H is that in the presence ofexternalities, it might be optimal to subsidize production. Thisis true if there are positive externalities associated with hogproduction. However, negative externalities are more likely thanpositive ones for the case of hog production in light of thepotential social costs of pollution.

Finally, the allusion to strategic trade intervention on p. 110needs to be backed up. It is unlikely that this game theoreticargument applies in this case. There is nothing in the P&Hreport to document strategic interactions between Canadian andAustralian firms. Blocking imports is not likely to confer astrategic advantage to Australian firms (in a game theoreticalsense), likewise for production subsidies. Such measures would

7


undoudtedly be appreciated by certain groups but they would leadto a net reduction in welfare for Australia.

3. AN OVERVIEW OF THE P&H REPORT:WHAT IT DOES AND WHAT IT SHOULD HAVE DONE

The P&H report is made up of seven sections: an executivesummary, an introduction, a review of the 1995 IndustryCommission Inquiry, two sections on econometric modeling, areview of possible assistance measures and a concluding section.If we go back to the necessary conditions for a safeguard action,it must be demonstrated: 1) that imports of a product areincreasing; and 2) that domestic producers of competitiveproducts are seriously injured or threatened with serious injury,and that this injury or threat is caused by the increase inimports. As implied by the title of their report (i.e., “TheImpact of Trade Liberalization and Increasing Imports onAustralian Pig Prices”), Purcell and Harrision (P&H) produced astudy narrowly focused on prices. Having gone in that direction,they cannot provide estimates about the degree of injurysuffered. Their analysis should have been extended to measurethe welfare effects of pork imports on Australia’s hog/porkindustry. This would have been a much more ambitious endeavorbecause it requires detailed modeling of all the components ofthe industry (retail, pork processing, hog production and feedsupply). The authors failed to analyze the consequences ofchanges in imports, hog prices and production and therefore didnot prove that serious injury was inflicted nor did they providesufficient empirical evidence to guide policymakers. Economistshave developed welfare measures that are well-suited to assessthe extent of the injuries or benefits stemming from imports.These measures can be used to compute the welfare changes allalong the marketing chain, from the consumers of meat productsall the way down to hog producers and their input suppliers. Thewelfare consequences of price, imports and production changes onthe whole industry are not and cannot be estimated with theirmethodology used by P&H. Keeping in mind that worries overabuses of Article XIX have led international trade experts likeHoeckman and Kostecki (1995, p.193) to recommend that the injurycriteria be such as to ensure that the domestic industry as awhole is making substantial losses, it can only be concluded thatthe evidence reported in the P&H is insufficient to justify asafeguard action.

In order to properly assess the extent of injury on the wholeindustry, one must gather information about the various elementsof the marketing chain. It is essential to develop anunderstanding of how the linkages between the various elements ofthe marketing chain work. Unfortunately, a careful reading ofthe P&H report does not provide much insight about the structureof the Australian pork industry. The descriptive analysis of theAustralian pork industry is lacking depth at all possible levels:feed supply, production, processing and retail. Basicinformation (e.g., the relative importance of the industry vis-à-vis other agricultural sectors, geographical concentration of thehog production and processing, industry concentration at the

8


processing and retail levels, competitiveness, cost structure atthe processing and retail levels…) would have been useful. Also,such information might have shed some light as to why the retailpork price behaves so differently from the saleyard price forbaconers. It is conceivable that the problem perceived by hogproducers might have more to do with domestic market failuresthan with import competition per say even though imports mightexacerbate the effects of the market failures. If that was thecase, government actions should be aimed at correcting thesemarket failures to make the marketing of hog/pork products inAustralia more efficient. In many countries, the food retail andhog processing sectors are highly concentrated. Australia isprobably no different but the P&H report does not have much tosay about this. If retailers exercise market power in theirdealing with local pork suppliers, then the opportunity thatretailers have to buy from foreigners contributes to the erosionof the bargaining power of local pork suppliers. However,restricting or even banning imports would not solve this domesticproblem.

It would have been insightful to get more detailed informationabout the composition of pork imports in terms of sources andproduct categories. As shown in Larue and Gervais (1996),changes in product composition may have strong effects on unitvalues/average price of imports. Furthermore, there is no reasonto believe that the demand at the retail and processing levels isthe same across product categories. There is a presumption thatthe substitution elasticity between Canadian and Australian porkproducts is high but not infinite in light of the persistence ofthe reported price differential. Knowing whether Australian andCanadian pork products are perfect or imperfect substitutes isgermane. If the degree of substitution is very low, Canadianpork would not be a threat.

There is no model showing how imports affect the variouscomponents of the Australian pork industry (feed supply,production, processing, retail). This could have been donegraphically or mathematically in a partial equilibrium framework.A theoretical model would have been useful for two reasons.First, it would have provided insights about the effect ofimports on prices at all levels of the marketing chain. Theimpact of elasticities and market structures could have beeneasily explored. In a sense, the theoretical model would haverefined our understanding of the Australian pork industry.Second, the structural theoretical equations could have beenpartially solved to derive a so-called reduced-form model of theindustry. The latter could have been estimated empirically andcomparative statics results, like the effect of a change in theCanadian price, could have been estimated. The advantage of thisapproach is that we would know how to interpret the coefficientsof the variables entering the specification of the reduced-formmodel, that is we would know if a coefficient is supposed to benegative, positive or if it has an ambiguous sign.

9


4. MODEL SPECIFICATION AND OTHER ECONOMETRIC PROBLEMS

The lack of a solid theoretical foundation is evident in thespecifications of the models estimated by P&H. For example, the‘model of supply in price dependant form’ of p.48 is not properlyspecified. The specification is supposed to explain hog supplyat the farm level. In a competitive setting (i.e., producersdon’t have market power), the supply function is obtained throughHotelling’s lemma, that is the first derivative of the profitfunction with respect to price. Let the profit function berepresented by π(p,w) where p is a vector of output prices and wa vector of input prices. The first derivative of the profitfunction of hog producers with respect to the price of hogs p i is∂π(p,w)/ ∂pi=qi(p,w) the supply function. This function can beinverted to be expressed in a price dependant form to give:pi=f(p -i, w, q i), where p -i is a vector of output prices other thanthe hog price. This contrasts with the implausible specificationused by P&H who failed to include input prices and wronglyincluded retail prices. What matters to hog producers is whatthey receive for their hogs and what they spend on inputs! Also,the aggregate supply curve alone does not explain the pricedetermination of Australian pork. In order to analyze theeffects of a change in the price of Canadian pork on Australianhog and pork production and prices, one needs also to estimatethe demand side. A fall in the price of Canadian pork shouldshift downward the demand for Australian pork. This fall in oneof the exogenous variables entering the demand equation moves theequilibrium along the supply curve. Thus, the resulting changein the Australian pork price cannot be estimated without knowinghow the demand for Australian pork reacts to a change in theprice of Canadian pork!

The incorrect specifications are not only difficult to interpretbut they also lead to biases in the econometric estimations.This is the so-called missing variables bias which does not goaway even if when the sample size for the regression analysisapproaches infinity.

It could that the model described and estimated in p. 48-49 isnot a supply function (even though P&H call it this way, they mayinterpret it differently). Perhaps their regression was meant tocapture both supply and demand functions. If that was the case,then it is obvious that they have not properly estimated theirmodel. Regardless of the interpretation given to the regression,the Chow test for structural change is improper,. The assumptionthat only SPQ, the saleyard price for baconers, is endogenous isunrealistic and the fact that the variables are I(1) warrants theuse of a more complex estimation technique and test forstructural change. The problem is one ofsimultaneity/endogeneity which introduces a bias in theestimation of the coefficients. P&H could have used Hansen’s(1992) fully modified approach to regressions with I(1) processesto address this issue. This way, they could have estimated thatregression (or possibly one with a proper set of explanatoryvariables) and the endogeneity of the variables on the right hand

10


side would have been corrected for. Hansen also provides threetests of structural change consistent with the fully modifiedestimator. For applications of this procedure in a context ofprice analysis, see Sabuhoro and Larue (1997). The FM system’sapproach developed by Peter Phillips (1995) would have been agood alternative as well.

Some of the arguments presented above also apply to theregressions reported on p.55-56. The biggest problem thoughremains the lack of a solid theoretical foundation to justify thespecification of these models. If the model estimated is againa supply function, as claimed by P&H, then several variablesincluded in the model are inconsistent with economic theory. Forexample, the price of Canadian pork should enter the demandspecification for Australian pork, not the supply function! Aglance at the results reveals that the coefficient on the priceof Canadian pork is not significant (p.55-56). Does this meanthat the prices of Canadian and Australian pork are unrelated?If so, how could Canadian imports injure Australian porkprocessors and hog producers? The inclusion of Canadian importsin P&H’ model specification is also puzzling. This variablecould enter the demand equation for Australian pork if there hadbeen quantitative restrictions on Canadian imports but P&H nevermention that an import quota (or a VER) was limiting porkimports. Nevertheless, let’s assume that this was the case.Then, why is the price of Canadian pork included? An alternativeis to assume that P&H really estimated a supply function and thatthat Canadian pork is used as a fixed input in the productionprocess of Australian pork (i.e., some sort of intermediategood)? That does not make much sense.

The confidence intervals reported on p.56 are not computed at thestandard level of significance, that is 5% (mean ± 1.96 SE).Using 1 SE instead of 1.96 SE makes P&H’ quota on Canadianimports appear more damaging because the lower bound of theinterval is then above zero.

The results of the VAR analysis in section 5 are strange. Itcould be because the VAR/VEC analysis is performed over a veryshort sample period (i.e, 30 observations/7years). Johansen’scointegration tests (Trace and Maximal eigenvalue) perform verypoorly in small samples (see Hargreaves (1995) Monte Carloresults). The short sample period is problematic for the VAR aswell because it imposes severe restrictions on the number of lagsused in the specification of the model. Even though P&H failedto put a confidence interval around their impulse responses, itlooks like the only variables that responds to shocks inAustralian variables are the Canadian variables! For example,the impulse functions on p. 67 show that only the Canadian volumeof imports react to a change in the price of Australian pork.The strange thing is that an increase in the Australian pricebrings about a large sudden decline in the volume of imports.Are Australian and Canadian pork complements? The policyimplications of such a conclusion are rather interesting in thecurrent context. On the other hand, the impulse functions on p.

11


77 show that none of the variables are responding to an increasein Canadian imports except Canadian imports. These resultscontradict those on p. 55-56 where Canadian imports appear tohave a very large effect on the Australian pork price. As in p.55-56, the impulse functions of the VAR model show that theeffect of the Canadian pork price on the Australian variables isessentially insignificant. If Canadian variables do not triggerany adjustments in Australian variables, how could they causeinjury?

The interpretation of unit roots in the P&H report is too strong.Hamilton (1994, p.445) argue that for any unit root process thereexists a stationary process that will be impossible todistinguish from the unit root representation for any givensample size T. Also, unit root and stationary processes differin their implications at infinite time horizons but for any givenfinite number of observations on the time series, there is arepresentative form of either class of models that could accountfor all the observed features of the data. Hence testing whetherinnovations have a permanent effect on the level of a series,however interesting, is simply impossible to do.

P&H discuss at great length the number of lagged variables toinclude in the specification of their models. For some modelsthey rely on the Akaike Information Criterion (AIC) while forothers they use Pierre Perron’s approach to compare a model of klags to a model with k-1 lags starting with an arbitrary large k.Both approaches are fine but may suggest widely differentspecifications. Hall (1994), who favors the use of Informationcriteria, recommend using the Schwartz’s criterion. My point isthat unit root and cointegration tests are often very sensitiveto the number of lags selected and both procedures could havebeen used to determine a range of lags to use. This range couldthen have been used to assess the robustness of the unitroot/cointegration results. This strategy is common, see Perron(1994) for example. By the same token, unit root/cointegrationtests are also sensitive to non-linear data transformation.Frances and Koop (1998) argue that the usual practice of takinglogs is not always warranted.

My experience with testing for unit roots is that differenttesting procedures often generates different results. Hence Iusually do more than one test. For unit roots at the zerofrequency (i.e., non-seasonal unit roots), I often run AugmentedDickey Fuller tests and the KPSS test developed by Kwiatkowskiet.al., (1992). The reason for this is that the null hypothesisunder the first test is that of a unit root or non-stationarity.In order to reject at say 5%, you need tremendous evidence andthe probability of wrongly confirming the presence of a unit rootwhere there is none is real in small samples. The second testhas a null of stationarity. If I can’t reject the null of a unitroot with a ADF test and that I can reject the null ofstationarity with the KPSS test, I can then be confident that myseries is non-stationary. If a series is potentiallycharacterized by a change in trend, then the tests proposed byPerron or Zivot and Andrews (1992) are useful. Some of these

12


tests allow the date at which the structural break occurs to beendogenously determined. It is not clear that this is the casein the P&H analysis. This is an important point because eventsare sometime anticipated and they may, on other occasions, inducedelayed adjustments. Hence fixing a date is usually not theright thing to do. The HEGY test for seasonal unit roots is themost popular of its kind. It is not very hard to implement andit allows for seasonal unit roots at different frequencies. I donot understand why P&H discusses it but do not implement it.

CONCLUSION

Improvements from the WTO agreement on safeguards are very muchdependent on the interpretation of what constitutes seriousinjury and the level of rigor required to demonstrate a causallink between imports and the degree of injury. In my view, P&Hhave not used a framework that enables them to measure and henceestablish injury. This is so because they cannot measurevariations in welfare. Reputable trade experts recommend thatthe injury test be administered on the whole industry and thatsubstantial aggregate losses be necessary to justify safeguardmeasures. In this case, retailers probably benefit a great dealfrom the imports. P&H’ analysis did not estimate the size ofthese benefits. Their models are improperly specified andproduce results that cannot be interpreted in a meaningful way,at least from an economic standpoint. They relied on essentiallythree types of models in their attempt to demonstrate injury.Yet, the single equation supply models of p.55-56 showed thatCanadian and Australian prices are independent from each other.However, these models indicated that Canadian imports had anegative effect on the Australian pork price. This resultcontrasts with the impulse responses derived from the VAR model.The impulse functions show that the Australian variables do notrespond to shocks on Canadian variables. In short, the empiricalevidence presented in the P&H report is mixed and of poorquality. It is my opinion that it is insufficient to establishserious injury.

13


References

Choi, K. and H.E. Lapan, "Optimal Trade Policies for a Developing Countryunder Uncertainty", Journal of Development Economics, 35,2, 1991, 243-60.

Corden, W.M., "The Normative Theory of International Trade", in Jones, R.and P.B. Kenen, eds., Handbook of International Economics vol. 1,Amsterdam: North Holland, 1984.

Deardorff, A.V., "Safeguards Policy and the Conservative Social WelfareFunction", in Stern, R.M., The Multilateral Trading System-Analysisand Options for Change, Ann Arbor: The Univesity of Michigan Press,1993.

Dixit, A. and V.K. Norman, Theory of International Trade, Cambridge:Cambridge University Press, 1980.

Finger, J.M., "Legalized Backsliding:safeguard provisions in GATT", inMartin, W. and L.A. Winters, eds., The Uruguay Round and theDeveloping Countries, Cambridge: Cambridge University Press, 1996.

Frances, P.H. and G. Koop, "On the Sensitivity of Unit Root Inference toNonlinear Data Transformations", Economic Letters, 59,1998, 7-15.

Hall, A., "Testing for a Unit Root in Time Series with Pretest Data-BasedModel Selection", Journal of Business & Economic Statistics, Oct.,1994, 461-470.

Hansen, B., "Test for Parameter Instability in Regressions with I(1)Processes", Journal of Business and Economic Statistics, July, 1992,321-335.

Hamilton, J.D, Time Series Analysis, Princeton: Princeton University Press, 1994.

Hargreaves, C., "A Review of Methods of Estimating CointegratingRelationships", in Hargreaves, C., ed., Non-Stationary Time SeriesAnalysis and Cointegration, New York: Oxford University Press,1995.

Hoekman, B.M. and M.P. Leidy., "Policy Responses to Shifting ComparativeAdvantage: Designing a System of Emergency Protection", in Stern,R.M., The Multilateral Trading System-Analysis and Options for Change,Ann Arbor: The Univesity of Michigan Press, 1993.

Hoekman, B.M. and M.M. Kostecki, The Political Economy of the World TradingSystem-From GATT to WTO, Oxford: Oxford University Press, 1995.

Jackson, J.H., The World Trading System: Law and Policy of InternationalEconomic Relations, 2nd ed., Cambridge: The MIT Press, 1997.

Johnson, H.G. "Optimum Welfare and Maximum Revenue Tariffs", Review ofEconomic Studies, 19, 1951, 28-35.

Kwiatkowski, D., P.C.B. Phillips, P. Schmidt, and U. Shin, "Testing theNull Hypothesis of Stationarity Against the Alternative of a UnitRoot: How Sure Are We that Economic Time Series Have a Unit Root",Journal of Econometrics, Oct., 1992, 159-178.

Larue, B., and J.-P. Gervais, "The Impact of the Canada-U.S. Pork TradeDispute on the Composition of U.S. Pork Imports" Agricultural andResource Economics Review, Oct., 1996, 187-199.

14


Mayer, W., "Optimal Pursuit of Safeguard Actions over Time" in Deardorff,A.V. and R.M. Stern, eds., Analytical and Negotiating Issues in theGlobal Trading System, Ann Arbor: The University of Michigan Press,1994.

Mussa, M. "Government Policy and the Adjustment Process", in Bhagwati, J.N.,Import Competition and Response, Chicago: The Univesity of ChicagoPress, 1982.

Perron, P. “Trend, Unit Root and Structural Change in Macroeconomic TimeSeries” in B. Bhaskara Rao, ed., Cointegration for the AppliedEconomist. New York: St-Martin’s Press.

Phillips, P.C.B. ‘Fully Modified Least Squares and Vector Autoregression’Econometrica, 63, 5, 1995, 1023-1078.

Sabuhoro, J.B. and B. Larue. "The Market Efficiency Hypothesis: The Caseof Coffee and Cocoa Futures". Agricultural Economics, 16,3,1997,171-184.

Sampson, G. "Safeguards", in Finger, J.M. and A. Olechowski, eds., TheUruguay Round: A Handbook for the Multilateral Trade Negotiations.Washingtion D.C.: World Bank, 1987.

Samuelson, P. "The Gains from International Trade", Canadian Journal ofEconomics and Political Science, 5, 1939, 195-205.

Vousden, N., The Economics of Trade Protection, Cambridge: CambridgeUniversity Press, 1990.

Zivot, E. and D.W.K. Andrews, " Further Evidence on the Great Crash, theOil-Price Shock,and the Unit Root Hypothesis", Journal of Businessand Economic Statistics, 10, 3,July 1992, 251-270.

In the Productivity Commission · Freehill Hollingdale & Page SYDCC\SUB34.DOC25 September 1998...

Documents

Transcript of In the Productivity Commission · Freehill Hollingdale & Page SYDCC\SUB34.DOC25 September 1998...